Shakeel

fuuuuuuuuuuuuck

when some people on the left are like “ya know, i’ve come to really respect tucker carlson, steve bannon, and marjorie taylor greene because they sorta share one of my anxieties or grievances, and i’m kinda surprised” i just think: i am. not!

Vaguepost

We support Illinois SB 315, which builds on California’s and New York’s landmark AI safety and transparency laws by establishing strong protections to safeguard kids, users, and communities. While we do support SB 315’s third-party audit provisions in principle, we have potential concerns about the implications of how the bill addresses them. We believe third-party audits of frontier models will be important, if implemented properly and professionally, but if not, they carry the risk of becoming politicized or ineffective due to being run through bureaucracies that don’t have sufficient AI expertise. We worry that if many states include such provisions, then AI models will be subject to many disparate and overlapping audits, similar to how money transmitter licensing requires audits from 53 jurisdictions. Furthermore, we think it’s important that states remain harmonized rather than having an arms race of escalating provisions. We won’t weigh in on every piece of AI legislation, but wanted to do so here in order to better communicate the nuances of our AI position. We continue to believe a comprehensive federal framework for frontier AI oversight will be critical so there is one set of rules that provides safety, oversight, and protection for innovation, but that harmonized state legislation (as we’re seeing in California, New York, and now Illinois) is helping shape an emerging national model for AI governance. We appreciate the strong partnership and collaboration among all stakeholders on these issues during the final weeks of session.

Joseph Gordon-Levitt visited Utah’s Capitol today to push lawmakers to back HB286, a bill that would require AI companies to strengthen protections for children and other users. The proposal cleared a House committee unanimously.

Interesting to see Douthat — a Catholic — make a very similar case to me:

We support Illinois SB 315, which builds on California’s and New York’s landmark AI safety and transparency laws by establishing strong protections to safeguard kids, users, and communities. While we do support SB 315’s third-party audit provisions in principle, we have potential concerns about the implications of how the bill addresses them. We believe third-party audits of frontier models will be important, if implemented properly and professionally, but if not, they carry the risk of becoming politicized or ineffective due to being run through bureaucracies that don’t have sufficient AI expertise. We worry that if many states include such provisions, then AI models will be subject to many disparate and overlapping audits, similar to how money transmitter licensing requires audits from 53 jurisdictions. Furthermore, we think it’s important that states remain harmonized rather than having an arms race of escalating provisions. We won’t weigh in on every piece of AI legislation, but wanted to do so here in order to better communicate the nuances of our AI position. We continue to believe a comprehensive federal framework for frontier AI oversight will be critical so there is one set of rules that provides safety, oversight, and protection for innovation, but that harmonized state legislation (as we’re seeing in California, New York, and now Illinois) is helping shape an emerging national model for AI governance. We appreciate the strong partnership and collaboration among all stakeholders on these issues during the final weeks of session.